Although large amounts of resources have been dedicated to the design of programs to persuade people to use condoms, little if any work has been done on the characteristics of messages and interventions that increase participation in these programs. In fact, programs are typically tested in conditions that reduce attrition and provide strong incentives for participation. The proposed research examines selective exposure to HIV-prevention interventions from a model that takes into account past behavior, prior motivations, perceptions of usefulness of the communication, and confidence estimates. The first phase of the research includes a metaanalysis examining participation and attrition as indicants of selective participation and retention in preventive programs. The second phase of this research consists of two field studies in which HIV-related communications are available to the participants, but participants are free to approach or avoid the persuasive messages. Because exposure is not forced but instead observed in relatively natural conditions, the study will examine selective participation and retention in the field and help to identify methods to increase participation in HIV prevention.